The goal of the proposed research is to develop a method for significantly improving the performance of sequencing by hybridization (SBH), thus decreasing the cost and increasing the speed of sequencing DNA. If successful, the technology that GeneSpectrum Inc. is developing will reduce the cost of such sequencing by 100 fold. The limitations of current DNA sequencing technology and the growing appetite for DNA sequence information define a need for a rapid, inexpensive method for DNA sequence determination, particularly in applications to human health. Sequencing by hybridization was proposed as an efficient method for sequencing DNA in the late 1980s and early 1990s. The method proved to be inefficient and was essentially abandoned. We have discovered a novel approach to the design and interpretation of arrays that provides a large increase in the amount of information that can be extracted from the array and that will make SBH a viable technique. Our method relies on arrays of oligonucleotide probes that contain natural bases and universal bases. In the context of this proposal universal bases are defined as nucleotide analogs that, when incorporated into DNA, hybridize equally well with all four of the cognate bases. The universal bases are used to arrange the natural bases into patterns in the oligonucleotide probes. Arrays of these patterned oligonucleotides allow us to use unique algorithms that result in highly efficient sequence reconstruction. In the last two years, we have investigated the ability of several molecules to function as the universal base in our patterned probes. We have recently discovered a method to make surrogates of universal bases by using specific mixtures of natural bases, which we refer to as "semi-degenerate" bases. The result is the delineation of a method to form patterned probes that can be tested and used immediately. The objective of the project proposed here is to demonstrate that semi-degenerate bases will function in arrayed, patterned probes as universal bases.